Fuzzy Control-based Three-Dimensional Motion Planning of an Amphibious Spherical Robot

Liang Zheng1, 2, 3, Shuxiang Guo*4, *5 Yan Piao*3, Ruochen An1, Wenbo Sui1

1 Graduate School of Engineering, Kagawa University, 4 Department of Intelligent Mechanical Systems Engineering, Takamatsu, Kagawa 761-0396, Japan Kagawa University, Takamatsu, Kagawa 761-0396, Japan

2 Jilin Agricultural Science and Technology University 5 Key Laboratory of Convergence Medical Engineering and Jilin, Jilin, China, 132101 System and Healthcare Technology, the Ministry of Industry

3 Changchun University of Science and Technology, Information Technology, School of Life Science, Beijing Changchun, Jilin, China, 130022 Institute of Technology, Haidian District,

Beijing 100081, China s18d505@stu.kagawa-u.ac.jp guo@eng.kagawa-u.ac.jp; piaoyan66@126.com

Abstract - Attitude control adjustment of the Amphibious Spherical Robot (ASR) moving on-land is a relatively mature technology, but there are few related types of research on Multi-Degree-of-Freedom (MDOF) for underwater attitude control algorithm. This paper developed a novel structure to control the underwater motion posture and proposed a fuzzy algorithm to control the internal water tank level so that the robot can realize the function of floating and diving. The control algorithm is especially based on underwater environment and we also take a detailed mechanical analysis of the water tank structure, the direction of the robot can be adjusted in real-time according to the water level with the fuzzy control algorithm and the purpose of the three dimensional underwater movement can be achieved. At the end of this paper, the underwater experiment is performed to analyze the structure and algorithm to give the most reasonable suggestion for three-dimensional underwater movement.

Index Terms - Biological Inspiration, Amphibious Spherical

Robot, Underwater Robot, Fuzzy control, 3D Motion

I. INTRODUCTION

In recent years, more and more researchers have become interested in the research of bionic robots. The bionic robot can easily realize the structural design and motion planning control according to the characteristics of the natural attributes. Bionic robots can be used in multiple applications and explore such as unknown seabed detection, military reconnaissance, resource collection, and detection of narrow places [1]-[4]. Within the last few years great efforts have been made to create a variety of robotic amphibious robot with different structures and swimming abilities. Existing fully untethered amphibious robot includes urchin-inspired robot [5], climbing soft robot [6], and a multi amphibious spherical robot designed by Guo Lab [7]-[14].

To realize the long-term monitoring and exploration tasks of the underwater complex environment in the mineral-rich ocean, it is necessary to expand the environmental awareness type of amphibious robots and carry a variety of sensors, which are constrained by the compactness of the structure. Therefore, it is difficult to satisfy the exploration and monitoring of a wide

area environment with a single traditional small robot. Therefore, it is necessary to design a novel robot, which has a more optimized structure and carries more miniature sensors.

There are many types of research on MDOF underwater of bionic robots. [15] presented a three-dimensional path planning method that combines the gliding with dolphin-like motions for the gliding robotic dolphin, relying on the two wings, the robot can achieve MDOF underwater movement. [16] developed multiple homogenous under-actuated saucer-type autonomous underwater gliders subjected to unmeasured velocities, design and model uncertainties, as well as unknown environmental disturbances. An improved double PD control method was proposed in [17], this paper achieved a consistent control effect on different target depths. A novel concept designed of a multi-legged underwater manned seabed walking robot is presented in [18], which can make robot will be used in both shallow water current (1-2 m/sec) and deep water up to 500 m. [19] proposed a centroid vectoring for attitude control of floating base robots to derive the control algorithm, which allows to control the orientation of the main body by adjusting the control input to its actuators to achieve free movement. Several solutions for diving systems in a novel type of underwater robot named artificial mussel are investigated in [20]. In this paper, we design a novel type of ASR. On basis of the original structure, a drainage device that controls the MDOF movement in the underwater environment is designed, which can achieve goals of float and dive. This paper also proposed a method of influent and drainage control based on a fuzzy control algorithm. With the assistance of a carbon dioxide gas cylinder, influent and drainage control can be quickly achieved. Thus, the robot can achieve the MDOF underwater movement.

The rest of this paper is organized as follows. The introduction of mechanical and electronic design of the ASR in Section 2. In section 3, the fuzzy control modeling and the associated controller method about floating and diving are proposed. In Section 4, we did the most important related experiments about communication and three-dimensional underwater movement with fuzzy control strategies. Finally, conclusions and future work are presented in Section 5.

Fig. 1 Structure of the amphibious spherical robot.

(a) land posture (b) underwater posture

Fig. 2 Two modes of the ASR.

II. OVERALL DESIGN

A. Mechanical design Fig. 1 shows the structure of the ASR, including the upper

hemisphere, lower hemisphere, waterproof tank compartment (contain control circuit and carbon dioxide gas cylinder), visual sensor, communication module, ultrasonic sensor (measuring distance), four water jets, nine steering gears, and a drainage holes. The lower hemispherical shell can be folded into the upper hemisphere, and the direction of the water jet is controlled by the servo motor to realize the amphibious mode [21]-[27]. As shown in Fig. 2, there are including the land mode and the underwater mode. The land mode is similar to the bionic turtle, so a spherical shape is formed through external shape control. The spherical shape is necessary when the robot is moving underwater. The reason why the sphere is used underwater is that the sphere has multiple degrees of freedom, and the closed sphere can better integrated protect the internal components. The water jet can promote the movement and has better concealment to prevent the generation of greater noise.

This paper mainly introduces the drainage module which installed on the middle plate. As shown in Fig. 3, it is mainly divided into two parts in the waterproof tank [28]. The first part is the circuit control module, and the second part is the gas cylinder control device. The gas cylinder control is composed of three air valves and a carbon dioxide gas cylinder. As shown in Fig. 4, the main reason for using carbon dioxide gas is that it can quickly inflate and deflate so that the airbag located in the lower hemisphere can be quickly inflated and deflated. The

fuzzy control method is used to achieve the purpose of floating and diving.

At the bottom of the exhaust device, the purpose of installing a steering gear is to control the opening and closing of the lower exhaust valve, and the signal from the control board to the steering gear is used to control the opening and closing, it is worth noting that the time of exhaust and inflation is very important, using the rotation time of the servo to fixedly set the time of two rotations. The water pressure of the lower chamber is adjusted to control the water intake of the tank. According to the principle of Archimedes buoyancy. The dimension of the buoyancy is determined by the volume of water discharged from the sphere. When the robot dives, the waterproof steering gear rotates clockwise to close the deflation valve, and water enters through the hole at one end of the sphere. Due to the pressure, the gas inside the sphere is discharged from the other end of the sphere. The weight of the water entering the sphere plus the weight of the waterproof steering gear of the sphere itself makes the spherical robot dive. When ascending, the waterproof steering gear rotates counterclockwise to release the compressed carbon dioxide gas in the carbon dioxide gas cylinder. At this time, the previous exhaust and water inlet hole have become drainage, causing the robot to float.

Fig. 3 Interior layout of the waterproof tank.

Fig. 4. Structure of the float and dive module.

Fig. 5. Schematic of the circuit control.

B. Electronic Design As shown in Fig. 5, the spherical robot is powered by two

7.4 V lithium batteries. The main control adopts the STM32 (Arm® Cortex®-M) minimum system as the main controller. The sensors used in the main controller part include a six-axis attitude sensor, depth sensor, and communication module. The power of the main control part is provided by battery two. In the power part, the robot uses 4 underwater thrusters to provide power. The power of the underwater thrusters is provided by the lithium battery two. Underwater thrusters, the power supply, and the relay are connected in series. The relay is controlled by the digital signal of the STM32 control board. When the relay receives a high level, the node is closed, and the underwater thruster starts to work. In addition, the robot uses 8 servo motors to control the direction of the underwater propeller. STM32 controls the servo drive board through the serial port. The servo drive board converts the serial command into a Pulse Width Modulation (PWM) signal to control the rotation of the steering wheel of the servo motor, and then realize the thruster. For the direction control, the power supply of the steering gear and the steering gear driving board is separately provided by lithium of the battery one.

The gas cylinder and exhaust valve used for ascending and descending are controlled by a waterproof servo, the control signal of the servo is directly provided by the main controller, and the power is provided by the lithium battery one. When ascending, the steering wheel of the gas cylinder steering gear rotates, driving the valve of the gas cylinder to open, and at the same time, the exhaust hole valve servo is reset to zero, the gas cylinder releases the gas, and the exhaust hole is closed to achieve floating. When diving, the gas cylinder steering gear rotates in the opposite direction, and the vent valve rotates 90 degrees, the gas cylinder is closed, the vent is opened, and the gas is discharged to achieve the dive.

III. FUZZY CONTROL

Fuzzy control algorithm is mainly divided into three situations: if the current water level is higher than the target water level, the water is discharged outward, the greater the difference, the faster the drainage. If the current water level is lower than the target water level, the water is injected inward,

the greater the difference, the faster the water injection. If the difference between the current water level and the target water level is small, the drainage speed and water injection speed must be kept equal. By controlling the two switches of S1 (water inlet) and S2 (water outlet), the fuzzy control algorithm is used to control the water level, as shown in Fig. 6.

Generally, the deviation e, that is, the difference between the target water level and the current water level is selected as the observation quantity, and the valve opening u is selected as the control quantity. The final control quantity u that we need to obtain is the output of fuzzy control. u can be obtained by synthesizing the deviation matrix e and the fuzzy relation matrix R, which satisfies the following conditions,

u eR= . (1)

Divide the deviation e into five fuzzy sets, negative large (NB), negative small (NS), zero (ZO), positive small (PS), and positive large (PB), e is negative means that the current water level is lower than the target water level, e is positive means that the current water level is higher than the target water level. Set the value range of e to [-3, 3]. Similarly, the control quantity u is divided into 5 fuzzy sets, negative large (NB), negative small (NS), zero (ZO), positive small (PS), and positive large (PB), u is negative means increasing the inlet valve S1, the opening degree of u is positive means reducing the opening of the inlet valve S1 (while increasing the opening of the outlet valve S2). Set the value range of u to [-4, 4]. In the Matlab fuzzy control toolbox, we have realized the operation of 4-6 steps to solve the matrix operation, result is shown in Fig. 7.

Fig. 6. Principle of the float and dive module.

Fig. 7. Parameters of the fuzzy control algorithm.

Fig. 8 Images of the dive and float in heave motion.

Fig. 9 Experimental result obtained in underwater tests with 3D reference path and tracking results.

IV. EXPERIMENTAL RESULTS

To further validate the effectiveness of the fuzzy control for achieving underwater three-dimensional of the ASR. The experiments analysis under the environment of a square water tank that is 4000 mm in length, 2000 mm in width, and 1000

mm in height, the water depth is 900 mm. The experimental time-sharing process is shown in Fig. 8. Fig. 8 (a) to (e) show that the robot moves from 1-4 s diving into the water with a depth of about 500 mm. Then after 6 s to achieve the floating function. It can be seen from the figure that during the dive process, when t = 3 s, the center of gravity of the robot changes,

Robot

which affects the dive speed. In order to keep the center of gravity always balanced, a gyro sensor is installed at the center of the robot, so that the robot can maintain basic balance and stability, under the control of the stability module, the robot returns to the equilibrium state at t=6 s. More importantly, to make the robot collide when it descends to the bottom of the water tank, a depth detection module is installed on the bottom to protect integrity characteristic. Fig. 9 shows the 3D motion trajectory of two processes of ascending and descending. The starting point coordinates are (0.04, 0, 0.99) and the ending point coordinates are (0.38, 0.42, 0.98), due to the change of the center of gravity and the effect of the water flow speed, the robot does not return to the starting point after completing the cyclic movement. The x-axis deviation is 0.34 m, the y-axis direction deviation is 0.42 m, and the z-axis deviation is 0.01 m.

V. CONCLUSIONS

In this paper, mechanical structure and electronic control design are first introduced at the beginning of the paper, and the structure and control method for realizing three-dimensional underwater motion are developed. Then, this paper proposed a control method for the robot underwater motion based on fuzzy control and analyze the principle of control and evaluate the superiority of fuzzy control algorithm. At the end of this paper, using the underwater experiment to achieve 3D reference path and tracking results to get a trajectory, it is proved that the fuzzy control algorithm has a better effect on controlling the center of gravity.

Future work should further strengthen underwater MDOF motion control algorithm, and ultimately achieve the objectives of underwater multi-robot collaboration and optimal path planning.

ACKNOWLEDGMENT

This research is partly supported by National High Tech. Research and Development Program of China (No. 2015AA043202), and SPS KAKENHI Grant Number 15K2120.

REFERENCES

[1] J. Yu, J. Liu, Z. Wu, and H. Fang, “Depth control of a bioinspired robotic dolphin based on sliding-mode fuzzy control method,” IEEE Trans. Ind. Electron, vol. 65, no. 3, pp. 2429–2438, 2018.

[2] A. Zlatintsi, et al. “I-Support: A robotic platform of an assistive bathing robot for the elderly population,” Robotics and Autonomous Systems, 126, 103451, 2020.

[3] W. Wang, X. Dai, L. Li, B. Gheneti, Y. Ding, J. Yu, and G. Xi, “Three- dimensional modeling of a fin-actuated robotic fish with multimodal swimming,” IEEE/ASME Transactions on Mechatronics, vol. 23, no. 4, pp. 1641–1652, 2018.

[4] X. Liu, Z. Zhou, J. Mai, and Q. Wang, “Real-time mode recognition based assistive torque control of bionic knee exoskeleton for sit-to-stand and stand-to-sit transitions,” Robotics and Autonomous Systems, vol. 119, pp. 209-220, 2019.

[5] T. Paschal, Bell M A, J. Sperry, et al. “Design, Fabrication, and Characterization of an Untethered Amphibious Sea Urchin-Inspired Robot,” IEEE Robotics and Automation Letters, vol. 4, no. 4, pp. 3348-3354, 2019.

[6] Y. Tang, Q, Zhang, G. Lin, et al. “Switchable adhesion actuator for amphibious climbing soft robot,” Soft robotics, vol. 5, no. 5, pp. 592-600, 2018.

[7] L. Zheng, S. Guo, Y. Piao, et al. “Collaboration and Task Planning of Turtle-Inspired Multiple Amphibious Spherical Robots,” Micromachines, vol. 11, no. 1, 71, 2020.

[8] L. Zheng, S. Guo, and S. Gu, “The communication and stability evaluation of amphibious spherical robots,” Microsystem Technologies, vol. 25, no. 7, pp. 2625-2636, 2019.

[9] S. Gu, S. Guo, and L. Zheng, “A highly stable and efficient spherical underwater robot with hybrid propulsion devices,” Autonomous Robots, pp. 1-13, 2020.

[10] L. Zheng, S. Guo, Y. Piao, et al. “Study on cooperative control algorithm of two spherical amphibious robots,” in Proceedings of 2019 IEEE International Conference on Mechatronics and Automation (ICMA). IEEE, pp. 76-80, 2019.

[11] X. Hou, S. Guo, L. Shi, H. Xing, Y. Liu, H. Liu, and Z. Li, “Hydrodynamic analysis-based modeling and experimental verification of a new water-jet thruster for an amphibious spherical robot,” Sensors, vol. 19, no. 2, 259, 2019.

[12] H. Xing, S. Guo, L. Shi, X. Hou, Y. Liu, and H. Liu, “Design, modeling and experimental evaluation of a legged, multi-vectored water-jet composite driving mechanism for an amphibious spherical robot,” Microsystem Technologies, vol. 26, no. 2, pp. 475-487, 2020.

[13] H. Xing, L. Shi, K. Tang, et al. “Robust RGB-D Camera and IMU Fusion-based Cooperative and Relative Close-range Localization for Multiple Turtle-inspired Amphibious Spherical Robots,” Journal of Bionic Engineering, vol. 16, no. 3, pp. 442-454, 2019.

[14] H. Xing, S. Guo, L. Shi, et al. “Hybrid locomotion evaluation for a novel amphibious spherical robot, Applied sciences,” vol. 8, no. 2, 156, 2018.

[15] J. Wang, Z. Wu, M. Tan, and J. Yu, “3-D Path Planning with Multiple Motions for a Gliding Robotic Dolphin,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2019.

[16] H. Wang, L. Liu, D. Wang, and T. Wang, “Output-Feedback Control for Cooperative Diving of Saucer-Type Underwater Gliders Based on a Fuzzy Observer and Event-Triggered Communication,” IEEE Access, vol. 7, pp. 50453-50465, 2019.

[17] W. Mu, Z. Zou, G. Liu, Y. Yang, and L. Shi, “Depth Control Method of Profiling Float Based on an Improved Double PD Controller,” IEEE Access, vol. 7, pp. 43258-43268, 2019.

[18] A. Khan, W. Liquan, W. Gang, M. Imran, et al. “Concept Design of the Underwater Manned Seabed Walking Robot,” Journal of Marine Science and Engineering, vol. 7, no. 10, 366, 2019.

[19] T. Haus, M. Orsag, A. Nunez, et al. “Centroid vectoring for attitude control of floating base robots: from maritime to aerial applications,” IEEE access, vol. 7, pp. 16021-16031, 2019.

[20] E. Donati, van Vuuren, G. J, K. Tanaka, D. Romano, T. Schmickl, and C. Stefanini, “Amussels: diving and anchoring in a new bio-inspired under-actuated robot class for long-term environmental exploration and monitoring,” Annual Conference Towards Autonomous Robotic Systems, pp. 300-314, 2017.

[21] S. Gu, and S. Guo, “Performance evaluation of a novel propulsion system for the spherical underwater robot (SURIII),” Applied Sciences, vol. 7, no. 11, 1196, 2017.

[22] Y. Li, S. Guo, and Y. Wang, “Design and characteristics evaluation of a novel spherical underwater robot. Robotics and Autonomous Systems,” vol. 94, pp. 61-74, 2017.

[23] S. Guo, S. Pan, L. Shi, P. Guo, Y. He, and K. Tang, “Visual detection and tracking system for an amphibious spherical robot,” Sensors, vol. 17, no. 4, pp. 1-21, 2017.

[24] S. Guo, Y. He, L. Shi, S. Pan, K. Tang, R. Xiao, and P. Guo, “Modal and fatigue analysis of critical components of an amphibious spherical robot,” Microsystem Technologies, vol. 23, no. 6, pp. 2233-2247, 2017.

[25] L. Zheng, Y. Piao, Y. Ma Y, “Development and control of articulated amphibious spherical robot,” Microsystem Technologies, pp. 1-9, 2019.

[26] M. Li, S. Guo, H. Hirata, H. Ishihara, “A roller-skating/walking mode-based amphibious robot” Robotics and Computer-Integrated Manufacturing, vol. 44, pp. 17-29, 2017.

[27] J Guo, C Li and S Guo, “Path Optimization Method for the Spherical Underwater Robot in Unknown Environment,” Journal of Bionic Engineering, vol. 15, no. 1, 2018.

[28] L. Zheng, S. Guo, and S. Gu, “Structure Improvement and Stability for an Amphibious Spherical Robot”, in Proceedings of 2018 IEEE International Conference on Mechatronics and Automation, pp. 1586-1590, 2018.